Spatial Domain
Spatial domain research focuses on understanding and leveraging spatial information within various data types, aiming to improve model performance and extract meaningful insights. Current research emphasizes the integration of spatial information with other modalities (temporal, semantic) using architectures like transformers, graph neural networks, and diffusion models, often incorporating attention mechanisms to enhance feature extraction and modeling of complex relationships. This work has significant implications across diverse fields, from improving image and video processing and analysis to enhancing autonomous navigation, medical image analysis, and urban planning through more accurate and efficient algorithms.
Papers
Graph and Skipped Transformer: Exploiting Spatial and Temporal Modeling Capacities for Efficient 3D Human Pose Estimation
Mengmeng Cui, Kunbo Zhang, Zhenan Sun
PosMLP-Video: Spatial and Temporal Relative Position Encoding for Efficient Video Recognition
Yanbin Hao, Diansong Zhou, Zhicai Wang, Chong-Wah Ngo, Meng Wang